Project Summary Late adolescence (approximately ages 15 to 19) is a developmental period marked by interconnected changes in cognitive and neurobiological functioning, including reorganization of large-scale functional networks and improved ability to regulate attention and pursue rewards. Adolescence is also a period characterized by heightened stress and increased symptoms of mood disorders (MD), which are in turn associated with neurocognitive abnormalities in the same brain networks and cognitive domains that are highly plastic in adolescence. These developmental convergences suggest a model in which abnormalities in key neurocognitive dimensions predispose teens to MD, possibly because neurocognitive impairment impedes healthy stress coping behavior. However, shared risk factors and similarities in early-stage symptomatology of various MD has made it challenging to determine the specific pathways by which neurocognitive abnormalities contribute to MD. The proposed study will address these challenges with a multi-modal, longitudinal evaluation of risk ?biotypes? (defined by neurocognitive functioning), stress responses, and trajectories of MD symptoms in adolescence. We will recruit a sample of n=130 adolescents ages 15-19 (n=96 with high familial risk of MD) to participate in a 24- month study consisting of neuroimaging, cognitive testing, and symptom assessment (at baseline and 12 months), and mobile-app-based or telephone follow-up evaluations (once per 6 months) of stress events/responses and symptoms. We will apply latent variable analysis to the set of cognitive and neuroimaging measures to evaluate individual differences in two key neurocognitive dimensions: (1) cognitive regulation (CR), defined by cognitive control task performance and frontoinsular network functioning, and (2) reward sensitivity (RS), defined by behavioral response to reward and corticostriatal network functioning. Multi-level mediation models will test the effects of baseline, or changes in, CR and RS on future mood symptoms as mediated by stress-reactive behaviors. Our first aim is to investigate longitudinal associations between neurocognitive dimensions and mood symptoms, testing the hypothesis that the interaction of CR and RS dimensions will produce risk biotypes in which teens showing lower CR together with lower RS will report increases in anhedonia over time, and teens showing lower CR together with higher RS will report increases in mania over time. Our second aim is to investigate mediating mechanisms linking neurocognitive dimensions to mood symptoms, testing the hypothesis that the lower CR/lower RS biotype will be associated with future anhedonia via stress- reactive behavioral withdrawal, and the lower CR/higher RS biotype will be associated with future mania via stress-reactive behavioral impulsivity. Ultimately, this work may be a crucial step towards the translational goal of using neurocognitive biotypes to differentially diagnose and predict MD.